discourse/app/services/external_upload_manager.rb
Martin Brennan e4350bb966
FEATURE: Direct S3 multipart uploads for backups (#14736)
This PR introduces a new `enable_experimental_backup_uploads` site setting (default false and hidden), which when enabled alongside `enable_direct_s3_uploads` will allow for direct S3 multipart uploads of backup .tar.gz files.

To make multipart external uploads work with both the S3BackupStore and the S3Store, I've had to move several methods out of S3Store and into S3Helper, including:

* presigned_url
* create_multipart
* abort_multipart
* complete_multipart
* presign_multipart_part
* list_multipart_parts

Then, S3Store and S3BackupStore either delegate directly to S3Helper or have their own special methods to call S3Helper for these methods. FileStore.temporary_upload_path has also removed its dependence on upload_path, and can now be used interchangeably between the stores. A similar change was made in the frontend as well, moving the multipart related JS code out of ComposerUppyUpload and into a mixin of its own, so it can also be used by UppyUploadMixin.

Some changes to ExternalUploadManager had to be made here as well. The backup direct uploads do not need an Upload record made for them in the database, so they can be moved to their final S3 resting place when completing the multipart upload.

This changeset is not perfect; it introduces some special cases in UploadController to handle backups that was previously in BackupController, because UploadController is where the multipart routes are located. A subsequent pull request will pull these routes into a module or some other sharing pattern, along with hooks, so the backup controller and the upload controller (and any future controllers that may need them) can include these routes in a nicer way.
2021-11-11 08:25:31 +10:00

197 lines
6.2 KiB
Ruby

# frozen_string_literal: true
class ExternalUploadManager
DOWNLOAD_LIMIT = 100.megabytes
SIZE_MISMATCH_BAN_MINUTES = 5
BAN_USER_REDIS_PREFIX = "ban_user_from_external_uploads_"
UPLOAD_TYPES_EXCLUDED_FROM_UPLOAD_PROMOTION = ["backup"].freeze
class ChecksumMismatchError < StandardError; end
class DownloadFailedError < StandardError; end
class CannotPromoteError < StandardError; end
class SizeMismatchError < StandardError; end
attr_reader :external_upload_stub
def self.ban_user_from_external_uploads!(user:, ban_minutes: 5)
Discourse.redis.setex("#{BAN_USER_REDIS_PREFIX}#{user.id}", ban_minutes.minutes.to_i, "1")
end
def self.user_banned?(user)
Discourse.redis.get("#{BAN_USER_REDIS_PREFIX}#{user.id}") == "1"
end
def self.create_direct_upload(current_user:, file_name:, file_size:, upload_type:, metadata: {})
store = store_for_upload_type(upload_type)
url = store.signed_url_for_temporary_upload(
file_name, metadata: metadata
)
key = store.s3_helper.path_from_url(url)
upload_stub = ExternalUploadStub.create!(
key: key,
created_by: current_user,
original_filename: file_name,
upload_type: upload_type,
filesize: file_size
)
{ url: url, key: key, unique_identifier: upload_stub.unique_identifier }
end
def self.create_direct_multipart_upload(
current_user:, file_name:, file_size:, upload_type:, metadata: {}
)
content_type = MiniMime.lookup_by_filename(file_name)&.content_type
store = store_for_upload_type(upload_type)
multipart_upload = store.create_multipart(
file_name, content_type, metadata: metadata
)
upload_stub = ExternalUploadStub.create!(
key: multipart_upload[:key],
created_by: current_user,
original_filename: file_name,
upload_type: upload_type,
external_upload_identifier: multipart_upload[:upload_id],
multipart: true,
filesize: file_size
)
{
external_upload_identifier: upload_stub.external_upload_identifier,
key: upload_stub.key,
unique_identifier: upload_stub.unique_identifier
}
end
def self.store_for_upload_type(upload_type)
if upload_type == "backup"
if !SiteSetting.enable_backups? || SiteSetting.backup_location != BackupLocationSiteSetting::S3
raise Discourse::InvalidAccess.new
end
BackupRestore::BackupStore.create
else
Discourse.store
end
end
def initialize(external_upload_stub, upload_create_opts = {})
@external_upload_stub = external_upload_stub
@upload_create_opts = upload_create_opts
@store = ExternalUploadManager.store_for_upload_type(external_upload_stub.upload_type)
end
def can_promote?
external_upload_stub.status == ExternalUploadStub.statuses[:created]
end
def transform!
raise CannotPromoteError if !can_promote?
external_upload_stub.update!(status: ExternalUploadStub.statuses[:uploaded])
# We require that the file size is specified ahead of time, and compare
# it here to make sure that people are not uploading excessively large
# files to the external provider. If this happens, the user will be banned
# from uploading to the external provider for N minutes.
if external_size != external_upload_stub.filesize
ExternalUploadManager.ban_user_from_external_uploads!(
user: external_upload_stub.created_by,
ban_minutes: SIZE_MISMATCH_BAN_MINUTES
)
raise SizeMismatchError.new("expected: #{external_upload_stub.filesize}, actual: #{external_size}")
end
if UPLOAD_TYPES_EXCLUDED_FROM_UPLOAD_PROMOTION.include?(external_upload_stub.upload_type)
move_to_final_destination
else
promote_to_upload
end
rescue
if !SiteSetting.enable_upload_debug_mode
# We don't need to do anything special to abort multipart uploads here,
# because at this point (calling promote_to_upload!), the multipart
# upload would already be complete.
@store.delete_file(external_upload_stub.key)
external_upload_stub.destroy!
else
external_upload_stub.update(status: ExternalUploadStub.statuses[:failed])
end
raise
end
private
def promote_to_upload
# This could be legitimately nil, if it's too big to download on the
# server, or it could have failed. To this end we set a should_download
# variable as well to check.
tempfile = nil
should_download = external_size < DOWNLOAD_LIMIT
if should_download
tempfile = download(external_upload_stub.key, external_upload_stub.upload_type)
raise DownloadFailedError if tempfile.blank?
actual_sha1 = Upload.generate_digest(tempfile)
if external_sha1 && external_sha1 != actual_sha1
raise ChecksumMismatchError
end
end
# TODO (martin): See if these additional opts will be needed
# - check if retain_hours is needed
opts = {
type: external_upload_stub.upload_type,
existing_external_upload_key: external_upload_stub.key,
external_upload_too_big: external_size > DOWNLOAD_LIMIT,
filesize: external_size
}.merge(@upload_create_opts)
UploadCreator.new(tempfile, external_upload_stub.original_filename, opts).create_for(
external_upload_stub.created_by_id
)
ensure
tempfile&.close!
end
def move_to_final_destination
content_type = MiniMime.lookup_by_filename(external_upload_stub.original_filename).content_type
@store.move_existing_stored_upload(
existing_external_upload_key: external_upload_stub.key,
original_filename: external_upload_stub.original_filename,
content_type: content_type
)
Struct.new(:errors).new([])
end
def external_stub_object
@external_stub_object ||= @store.object_from_path(external_upload_stub.key)
end
def external_etag
@external_etag ||= external_stub_object.etag
end
def external_size
@external_size ||= external_stub_object.size
end
def external_sha1
@external_sha1 ||= external_stub_object.metadata["sha1-checksum"]
end
def download(key, type)
url = @store.signed_url_for_path(external_upload_stub.key)
FileHelper.download(
url,
max_file_size: DOWNLOAD_LIMIT,
tmp_file_name: "discourse-upload-#{type}",
follow_redirect: true
)
end
end